🤖 AI Summary
This work proposes a lightweight TinyML-based anomaly sound detection system tailored for microcontrollers to address the latency, high energy consumption, and privacy concerns associated with cloud-centric processing in IoT acoustic sensing. By extracting Mel-frequency cepstral coefficients (MFCCs) directly on the edge device and deploying a compressed and optimized neural network classifier, the system achieves high-accuracy, low-power, real-time local anomaly detection for the first time. Evaluated on the UrbanSound8K dataset, the approach attains 91% accuracy and a balanced F1-score of 0.91, demonstrating a practical trade-off between privacy preservation, energy efficiency, and real-time performance. This solution offers a viable pathway for scalable deployment in resource-constrained IoT applications.
📝 Abstract
Tiny Machine Learning enables real-time, energy-efficient data processing directly on microcontrollers, making it ideal for Internet of Things sensor networks. This paper presents a compact TinyML pipeline for detecting anomalies in environmental sound within IoT sensor networks. Acoustic monitoring in IoT systems can enhance safety and context awareness, yet cloud-based processing introduces challenges related to latency, power usage, and privacy. Our pipeline addresses these issues by extracting Mel Frequency Cepstral Coefficients from sound signals and training a lightweight neural network classifier optimized for deployment on edge devices. The model was trained and evaluated using the UrbanSound8K dataset, achieving a test accuracy of 91% and balanced F1-scores of 0.91 across both normal and anomalous sound classes. These results demonstrate the feasibility and reliability of embedded acoustic anomaly detection for scalable and responsive IoT deployments.